Chapter 4 Asymmetric Ring-Closing Metathesis with Ruthenium Alkylidenes Bearing Chiral, Monodentate N-Heterocyclic Carbene Ligands

Home , Grignard reagent

60 Chapter 4 Asymmetric Ring-Closing Metathesis with Ruthenium Alkylidenes Bearing Chiral, Monodentate N-Heterocyclic Carbene Ligands

Introduction

Asymmetric olefin metathesis does not seem possible at first glance, because no sp3-hybridized carbons are formed during a metathesis reaction. Instead of creating a new sp3 carbon, asymmetric metathesis reactions form chiral compounds through either kinetic resolutions of racemates or desymmetrizations of achiral or meso compounds

(Scheme 4.1). When chiral metathesis catalysts are used, enantioenriched products can be generated.1 The kinetic resolutions ideally involve selective ring closing of one enantiomer of a chiral diene while leaving the other enantiomer untouched. Asymmetric ring-closing metathesis (ARCM) and asymmetric cross-metathesis (ACM) are intramolecular and intermolecular reactions that produce a chiral center through desymmetrizations of trienes or dienes, respectively. Asymmetric ring-opening/cross metathesis (AROCM) reactions create multiple chiral centers by desymmetrizing meso compounds.

O kinetic resolution O O

R R R

asymmetric ring-closing metathesis (ARCM) O O

OR OR1 asymmetric cross-metathesis (ACM) 1 R2 R2

R asymmetric ring-opening/cross metathesis (AROCM) 2 R 1 R2 R1 R1 R1

Scheme 4.1. Examples of asymmetric olefin metathesis reactions.

61 Most asymmetric olefin metathesis reactions have been catalyzed by chiral molybdenum complexes, including kinetic resolutions, ARCM, AROCM, and asymmetric ring-opening/ring-closing metathesis (ARORCM).1 No single chiral molybdenum alkylidene catalyst is efficient and selective in every asymmetric metathesis reaction. Therefore, the Schrock and Hoveyda groups have generated a library of catalysts and screened them to find the best one for a given transformation. For example, complex 1 catalyzes the formation of 5-membered ring 4 with high enantioselectivity and yield, but it is inefficient and less selective in generating the six-membered ring 6

(Scheme 4.2).2 On the other hand, catalyst 2 affords 6 in 98% yield and in >99% ee, but is almost completely inactive in the synthesis of 4.3 More than 30 chiral molybdenum complexes have been made by varying the imido, alkylidene, and bidentate phenoxide groups.1b

Me Me O N Me Me O 1 or 2 (5 mol %) Mo O Ph O benzene

3 4 1: 6 h, 22 °C; 99% ee; 83% yield 1 2: 18 h, 22 °C; --% ee; <5% yield

Si Si Ar O 1 or 2 (5 mol %) O N Me Me O benzene Mo Ph O O 5 6 1: 24 h, 22 °C; 85% ee; 20% yield Ar 2: 3 h, 60 °C; >99% ee; 98% yield Ar = 2,4,6-triisopropylphenyl 2

Scheme 4.2. Differences in selectivity and efficiency between chiral molybdenum catalysts.

Much like the parent achiral molybdenum catalysts, the chiral complexes are more sensitive to air, moisture, and a variety of common functional groups than most

62 ruthenium olefin metathesis catalysts and need to be handled and stored in an inert atmosphere.1b Molybdenum olefin metathesis catalysts are incompatible with carboxylic acids, ketones, aldehydes, most alcohols, and primary amines. Because the ruthenium catalysts are so tolerant and therefore user friendly, enantioselective variants would also be expected to find widespread use. Unfortunately, only a few examples of ruthenium- catalyzed asymmetric olefin metathesis exist. Two classes of chiral ruthenium metathesis catalyst have been explored (Figure 4.1): those containing monodentate N-heterocyclic carbenes (NHCs) with chirality in the backbone developed in the Grubbs group (7 and 8)4 and those containing chiral, bidentate NHC/binaphthyl ligands developed in the Hoveyda group (9 and 10).5 Complexes 9 and 10 catalyze AROCM in up to 98% ee,6 but they exhibit reduced reactivity and selectivity toward ARCM relative to catalysts 7 and 8.7

Ph Ph Ph Ph Ph Ph Me i-Pr N N N N N N N N X X X X Ru Me i-Pr O Ru Ru Ru O Ph Ph X X O O PCy3 PCy3 Ph Ph 7a (X = Cl) 8a (X = Cl) 9a (X = Cl) 10a (X = Cl) 7b (X = I) 8b (X = I) 9b (X = I) 10b (X = I)

Figure 4.1. Representative examples of chiral ruthenium olefin metathesis catalysts.

The initial study of ruthenium metathesis catalysts bearing chiral, monodentate

NHCs involved screening six catalysts and three substrates.4 Complexes with NHCs derived from 1,2-diaminocyclohexane exhibited low enantioselectivities (0%–13% ee), but those derived from 1,2-diphenylethylenediamine catalyzed the formation of a 2,5- dihydrofuran in up to 90% ee. Catalyst 8b was the most selective catalyst, and the three substrates it was reacted with are shown in Scheme 4.3. The enantioselectivity of the reaction is highly dependent on the olefin substitution and the type of halide on the

63 catalyst, with iodide typically affording the highest ee. These were the only substrates explored in this study.

O O catalyst (2.5–5 mol %) 8a (2.5 mol %): 23% ee; 96% conv.

CH2Cl2, 38 °C 8b (5 mol %): 39% ee; 20% conv. or 11 THF, NaI, 38 °C 12

O O catalyst (2.5–5 mol %) 8a (2.5 mol %): 12% ee; 97% conv.

CH2Cl2, 38 °C 8b (5 mol %): 35% ee; 90% conv. or 13 THF, NaI, 38 °C 14

O O catalyst (2.5–5 mol %) 8a (2.5 mol %): 35% ee; 82% conv.

CH2Cl2, 38 °C 8b (5 mol %): 90% ee; 82% conv. or 3 THF, NaI, 38 °C 4

Scheme 4.3. ARCM of three trienes catalyzed by chiral ruthenium benzylidene complexes.

After the initial examination of catalysts 8a and 8b in ARCM, the focus shifted toward catalyzing AROCM and ACM with catalysts 8a and 8b.8 Moderate enantioselectivities (12%–82%) were observed over a wide range of substrates for both

AROCM and ACM. Based on the results from the initial report,4 it was clear that the 1,2- diphenylethylenediamine-based catalysts were superior to those derived from 1,2- diaminocyclohexane. Additionally, increasing the size of the ortho substituent from methyl to isopropyl (7b to 8b) improved the enantioselectivity of the formation of 4 (85% ee with 7b and 90% ee with 8b). Therefore, new catalysts containing varying substitution around the N-bound aryl rings that were based on 8 were targeted in this study. These substituted catalysts were tested in AROCM and ACM,8 but the results presented in this chapter are focused on ARCM with ruthenium catalysts that have substitution in the meta positions of the N-bound aryl rings. Very few substrates were tested in the initial study (Scheme 4.3), so one of the goals of the research presented here was to explore the substrate scope of this family of meta-substituted catalysts. Based on

64 the success of 8 in ARCM, it was expected that these new variants would also be selective and efficient.

Results and Discussion9

Design and Synthesis of Chiral Ruthenium Catalysts. One of the reasons the N-bound aryl rings were modified is that it was relatively straightforward to introduce structural changes to them, and therefore many different catalysts could be made. Initially, catalysts with ortho substituents larger than isopropyl were the targeted complexes.

Attempts to increase the size of the ortho substituent to t-butyl were unsuccessful.

Although a catalyst with an ortho cyclohexyl group was made, enantiomeric excesses were low (18% ee for the formation of 4), and the catalyst was unstable. It was thought that the para position was too remote from the ruthenium center, and changes there would not have any effect on the enantioselectivities. Therefore, catalysts with substitution in the meta position of the N-bound aryl rings were made.

A modular synthesis was used to access the desired chiral, non-racemic NHCs.

Most of this work was done by Jacob Berlin, another graduate student in the group, and all of the experimental details (including the synthesis of the aryl bromides) have been published.9 Briefly, these complexes were all generated by the same strategy (Scheme

4.4). Two equivalents of an aryl bromide were coupled to commercially available

10 (1R,2R)-1,2-diphenylethylenediamine using a Pd2(dba)3/(±)-BINAP catalyst system, and the resulting chiral diamines were reacted with triethylorthoformate and NH4BF4 to afford

– dihydroimidazolium BF4 salts. These carbene precursors were then reacted with bis- phosphine complex 15 and potassium hexafluoro-t-butoxide to generate the desired chiral

65 olefin metathesis catalysts as dichlorides. The yields for the last step varied due to a challenging chromatographic purification of the complexes. The para methoxy group in complexes 17a and 17b was used as a synthetic handle during the aryl bromide synthesis and was not expected to effect the enantioselectivities. The dichloride complexes (16a–

18a) were all stable to air and moisture for at least 6 months in the solid state, and the diiodide variants (16b–18b) were generated in situ by the addition of 25 equivalents of sodium iodide to the analogous dichloride catalyst. They were never isolated.

Ph Ph Br Ph Ph Ph Ph NH4BF4 H N NH i-Pr i-Pr 2 2 R1 R3 HCO H R1 R NH HN 2 N N 3 . R R R2 R R R Pd2(dba)3 CHCl3 (5 mol %) 2 2 HC(OEt)3 2 1 3 i-Pr i-Pr — (±)-BINAP (12 mol %) R R 120 °C, 12 h R3 BF4 R R2 3 1 1 NaOt-Bu, toluene 2 equiv 52%–66% yield 90 °C, 48 h 2 steps

PCy3 Cl Ru Cl Ph PCy3 15 KOC(CF3)2CH3 toluene 60 °C, 6 h

Ph Ph Ph Ph Ph Ph i-Pr i-Pr i-Pr i-Pr t-Bu i-Pr N N N N N N MeO OMe i-Pr X i-Pr X i-Pr X Ru i-Pr Ru t-Bu i-Pr Ru X Ph X Ph X Ph PCy3 PCy3 PCy3 16a (X = Cl) 17a (X = Cl) 18a (X = Cl) NaI (xs) NaI (xs) NaI (xs) 16b (X = I) 17b (X = I) 18b (X = I) 35%–85% yield 35%–75% yield 30%–83% yield

Scheme 4.4. Synthesis of chiral ruthenium olefin metathesis catalysts.

Substrate Synthesis. Only three substrates were used in the initial study of ruthenium catalysts bearing chiral, monodentate NHCs.4 It was discovered that one triene (3) underwent ARCM in 90% ee, but no other compounds similar to 3 were tested. One of the major goals of the research presented here was to explore the substrate scope of this family of catalysts, so the first trienes that were made were derivatives of 3. The building

66 block for all of these substrates was alcohol 21, and early on in this work 21 was made from vinyl bromide 19 (Scheme 4.5). Compound 19 was commercially available, and 21 was obtained in high yield. As more substrates were made, larger quantities of 21 were needed. Unfortunately, vinyl bromide 19 was expensive (>$2000/mol), and so an alternative route to 21 was developed. When 2-butyne (22) ($295/mol) was treated with isobutylmagnesium bromide in the presence of a catalytic amount of titanocene dichloride, a vinyl Grignard reagent was generated in situ.11 Tiglic aldehyde (20) was added to the solution, and 21 was generated. Multigram quantities of alcohol 21 were made using this procedure.

1. Cp2TiCl2 (3 mol %)

OH Br 1. t-BuLi MgBr

2. 2. 22 O O 19 21 H H 20 20 –78 °C, Et2O Et2O, rt 90% yield 69%–86% yield

Scheme 4.5. Two approaches to the synthesis of alcohol 21.

With alcohol 21 in hand, many different substrates were prepared (Scheme 4.6).

Most of the alkyl ethers were made by simply generating the sodium alkoxide of 21, and reacting it with an electrophile. The low yields obtained in the synthesis of 23 and 24 were attributed to steric hindrance around the hydroxyl group in 21. Attempts to make 29 by alkylating 21 with 4-bromo-1-butene or 4-iodo-1-butene were completely unsuccessful. When the sodium alkoxide was used, elimination to form 1,3-butadiene occurred instead of the desired nucleophilic displacement. The route that finally led to 29 involved a chemoselective hydroboration with 9-BBN followed by oxidation of the alkyl borane to primary alcohol 28. The alcohol was oxidized to an aldehyde that was not

67 stable, so the crude material was subjected to a Wittig olefination to afford 29 in 29% yield over two steps. Silyl ethers 26 and 27 were both made using standard conditions.

When electron-withdrawing groups such as allylchloroformate and vinylacetyl chloride were used as electrophiles, elimination of the carbonate or vinylacetate, respectively, occurred (Scheme 4.7).

Br n = 1,3,4 or O O O O Si Cl Si Me Me

NaH, THF, 65–70 °C 3 23 24 25 OH 82% yield 22% yield 44% yield 65% yield

Me Me Ph Ph 21 Si Si R R O O Si Cl

Et3N, DMAP CH2Cl2, rt 26 27 76% yield 69% yield

1. (COCl)2 DMSO Et N O O OH 3 O 1. 9-BBN CH2Cl2, –78 °C

2. NaOH 2. [Ph3PMe]Br H2O2 BuLi 3 THF, rt–50 °C 28 THF, 0 °C 29 83% yield 29% yield over 2 steps

Scheme 4.6. Synthesis of ARCM substrates based on alcohol 21.

OH Cl O (2 equiv) O CO2 pyridine, 0 °C O O 21

Scheme 4.7. Elimination of a carbonate derivative of 21.

In addition to substrates based on allylic alcohol 21, trienes derived from homoallylic alcohol 33 were also made (Scheme 4.8). Compound 33 was synthesized in three steps from epichlorohydrin (30): the appropriate vinyl magnesium bromide was generated in situ,11 and in the presence of catalytic CuBr, it opened the epoxide. An

68 intramolecular nucleophilic displacement reaction transformed chlorohydrin 31 into epoxide 32, which underwent a second epoxide-opening process to afford 33. The allyl ether 34 and the dimethylallylsilyl ether 35 were formed using standard conditions, and they are analogous to compounds 3 and 26, respectively.

1. 1.

MgBr MgBr

Cp TiCl (cat.) NaH O 2 2 OH Cp2TiCl2 (cat.) OH Cl O 2. CuBr (cat.) Cl THF, reflux 2. CuBr (cat.) Et O, –40 °C 51% yield Et O, –40 °C 30 2 31 32 2 33 68% yield 60% yield

Br O

NaH, THF, 65–70 °C 73% yield 34

33 Me Me Me Si Me Cl Si O

Et3N, DMAP CH2Cl2, rt 35 88% yield

Scheme 4.8. Synthesis of ARCM substrates based on alcohol 33.

Efficiency and Enantioselectivity of Chiral Ruthenium Catalysts. All of the achiral trienes were treated with catalysts 8, 16, 17, and 18 in the absence and presence of sodium iodide. The results of the ARCM reactions of the alkyl ethers derived from 21 are shown in Table 4.1. First, the addition of sodium iodide to the reaction had an enormous impact on the enantioselectivity, regardless of what substrate or catalyst was used. The enantiomeric excesses of the cyclic products increased up to 57% relative to those with the dichloride catalysts when the diiodide catalysts were used, and they were all >80%. Of the dichloride catalysts, 18a was the most selective for all of the substrates in Table 4.1. As the meta substituent para to the isopropyl group increased in size

69 (catalysts 8a, 16a, and 17a), the enantiomeric excess of the product decreased, but never by more than 17% when going from H to t-butyl. With the diiodide catalysts, this trend did not continue, and the selectivities were similar for 8b, 16b, and 17b.

Table 4.1. ARCM of achiral, alkenyl ethers using chiral ruthenium olefin metathesis catalysts.

Substrate Product 8 16 17 18

O O 35% ee, >98% conv 31% ee, >98% conv 30% ee, >98% conv 46% ee, >98% conv NaI: NaI: NaI: NaI: 90% ee, >98% conv 84% ee, >98% conv 87% ee, >98% conv 90% ee, >98% conv 3 4

O O 68% ee, >98% conv 67% ee, >98% conv 64% ee, >98% conv 74% ee, >98% conv NaI: NaI: NaI: NaI: 90% ee, >98% conv 88% ee, >98% conv 82% ee, 43% conv 84% ee, 70% conv 29 36

O O 65% ee, >98% conv 56% ee, >98% conv 48% ee, 91% conv 76% ee, 93% conv NaI: NaI: NaI: NaI: 85% ee, 5% conv 87% ee, 51% conv 85% ee, 70% conv 83% ee, 20% conv 23 37

O O 65% ee, ~1% conv 57% ee, 4% conv 51% ee, 9% conv 71% ee, ~2% conv NaI: NaI: NaI: NaI: 85% ee, ~2% conv 88% ee, 5% conv 0% conv 0% conv 24 38 Conditions: dichloride catalyst (2 mol %), triene, CH2Cl2 (0.055 M), 40 °C, 2 h; or dichloride catalyst (4 mol %), NaI (25 equiv relative to catalyst), triene, THF (0.055 M), 40 °C, 2 h.

Conversions >90% were obtained in all of the reactions with the dichloride catalysts, except when 24 was used. The product of the RCM of 24 is an eight-membered ring containing a trisubstituted alkene (38), and these types of products are typically challenging to access using RCM.12 Even though the loadings of the diiodide catalysts were doubled relative to the dichloride catalysts, lower conversions were often observed.

This may be due to catalyst decomposition, as diiodide ruthenium metathesis catalysts are typically less stable than the corresponding dichloride catalyst.13 Reactions with

dichloride catalysts were performed in CH2Cl2 and reactions with diiodide catalysts were in THF, so the conversions may also be dependent on solvent.

70 The trienes derived from 21 bearing silyl alkenyl ethers were also treated with catalysts 8, 16, 17, and 18, and the results are shown in Table 4.2. For all of the substrates, the enantiomeric excess were ≥75% with the dichloride catalysts. The diiodide catalysts were typically more selective, but the differences were not nearly as large as with the substrates in Table 4.1. Additionally, as in Table 4.1, selectivity decreased as steric bulk in the meta position increased (8a, 16a, and 17a), but there was no simple trend with the analogous diiodide catalysts.

Table 4.2. ARCM of achiral, silyl alkenyl ethers using chiral ruthenium olefin metathesis catalysts.

Substrate Product 8 16 17 18

Si Si O O 83% ee, >98% conv 81% ee, >98% conv 75% ee, >98% conv 92% ee, >98% conv NaI: NaI: NaI: NaI: 86% ee, 68% conv 90% ee, >98% conv 85% ee, >98% conv 92% ee, 58% conv

26 39

Si Si O O 84% ee, 88% conv 80% ee, 91% conv 78% ee, 90% conv 92% ee, 93% conv NaI: NaI: NaI: NaI: 87% ee, 15% conv 90% ee, 75% conv 86% ee, 50% conv 92% ee, 10% conv

25 40

Ph Ph Ph Ph Si Si O O 77% ee, >98% conv 80% ee, >98% conv NaI: N/D N/D NaI: 83% ee, 96% conv N/D 27 41 Conditions: dichloride catalyst (2 mol %), triene, CH2Cl2 (0.055 M), 40 °C, 2 h; or dichloride catalyst (4 mol %), NaI (25 equiv relative to catalyst), triene, THF (0.055 M), 40 °C, 2 h. N/D = not determined.

The most exciting discovery was that 18a catalyzed the ARCM of 26 and 25 in

92% ee. Not only was this the highest enantiomeric excess obtained using this family of catalysts, but also it was achieved without the need for sodium iodide. Moreover, since the dichloride catalysts are generally more stable than the diiodide catalysts, lower catalyst loadings could be used. No difference in enantioselectivity and a decrease in conversion was observed when 18b was reacted with 26 and 25. It was thought that the success of these substrates was due in part to the methyl groups on the dimethylsilyl

71 linker increasing the difference in energies of the diastereomeric transition states, so a triene with two phenyl groups (27) was made. Unfortunately, the enantioselectivity of the reactions with 27 were lower than with 26 and 25.

In addition to the substrates based on allylic alcohol 21, the trienes derived from

33 were also reacted with 8, 16, 17, and 18 (Table 4.3). When both 34 and 35 underwent

ARCM, small amounts of the products lacking a terminal methyl group (43 and 45) were formed. The enantiomeric excesses of the methylated and demethylated products were generally similar, suggesting the methyl group was removed after the ring closing occurred. The matched or mismatched interaction of the catalysts with the chiral products could explain the increase or decrease in enantiomeric excess of the demethylated products relative to the methylated products observed in some cases

(compare 44 and 45 with catalyst 16b). The reactions with allyl ether 34 went to complete conversion with all of the catalysts, but the reactions were only moderately selective (up to 58% ee). Greater enantioselectivities were obtained with silyl ether 35, but reactions with the analogous substrates based on alcohol 21 (Table 4.2) were more selective under these conditions.

72 Table 4.3. ARCM of substrates derived from alcohol 33.

Substrate Product 8 16 17 18

O 31%ee, >95% conv 37%ee, >95% conv 23%ee, >95% conv 31%ee, >95% conv NaI: NaI: NaI: NaI: 35%ee, >95% conv 58%ee, >95% conv 44%ee, >95% conv N/D 42 O

O 29%ee, <5% conv 35%ee, <5% conv 28%ee, <5% conv 13%ee, >95% conv 34 NaI: NaI: NaI: NaI: 38%ee, <5% conv 54%ee, <5% conv 47%ee, <5% conv N/D 43

Si 56%ee, 91% conv 39%ee, 80% conv 36%ee, 95% conv 52%ee, >98% conv O NaI: NaI: NaI: NaI: 78%ee, >95% conv 74%ee, 92% conv 51%ee, 89% conv 77%ee, >98% conv Si 44 O

Si 32%ee, 9% conv 27%ee, 20% conv 30%ee, 5% conv 6%ee, <2% conv O 35 NaI: NaI: NaI: NaI: 61%ee, <5% conv 35%ee, 8 conv 22%ee, 11% conv not observed 45 Conditions: dichloride catalyst (2 mol %), triene, CH2Cl2 (0.055 M), 40 °C, 2 h; or dichloride catalyst (4 mol %), NaI (25 equiv relative to catalyst), triene, THF (0.055 M), 40 °C, 2 h. N/D = not determined.

The data in Tables 4.1 and 4.2 showed that the addition of meta substituents para to the ortho isopropyl group (catalysts 16 and 17) only caused minor fluctuations in the enantioselectivity of the reaction. The introduction of a meta isopropyl group ortho to the ortho isopropyl group (18) increased the selectivity of the catalyst. From these reactions it became clear that the parent complex 8b, and the diisopropyl variant 18a, were the most selective catalysts of those tested. Therefore, isolated yields were obtained from

ARCM reactions between 8b or 18a and selected achiral trienes (Table 4.4). The yields for 4 and 36 were moderately reduced due to the volatility of the products during purification. Attempts to use less than 4 mol % of 8b resulted in incomplete conversion of 3 to 4. When a more challenging substrate (23) was used, 8b formed the product in

85% ee, but in only 5% conversion. Catalyst 18a is much more efficient (presumably because it is more stable that 8b),13 and 37 was isolated in 92% yield and 76% ee.14

73 Compound 39 was isolated in 77% yield using almost 1 g of 26 with less than 1 mol % of

18a, and no decrease in enantiomeric excess occurred relative to the screening reactions.

Table 4.4. ARCM reactions of selected achiral trienes with chiral ruthenium catalysts.

Triene Product Catalyst (mol %) ee (%)a Conv. (%)b Yield (%)

O O 8b (4) 90 >98 64

3 4 O O

8b (4) 90 >98 77

29 36

O O 8b (4) 85 5 N/D

c 23 37 18a (2) 76 93 92

Si Si O O 18a (0.8) 92 >98 77d

26 39

O Si Si O 18a (1) 92 65 64

25 40

Si Si O O 8b (4) 78 >98 98

35 44 Conditions for reactions with 8b: NaI (25 equiv relative to catalyst) and 8a in THF (0.055 M in triene) for 1 h at rt, then add triene and stir for 2 h at 40 °C; conditions for reactions with 18a: a triene, CH2Cl2 (0.055 M in triene), and 18a for 2 h at 40 °C. Enantiomeric excesses determined by chiral GC. b Determined by 1H NMR spectrum of crude reaction mixture. c See reference 14. d Reaction done on a 4 mmol (0.95 g) of 26 scale. N/D = not determined.

All of the achiral trienes described above have two trisubstituted olefins with cis methyl groups. These substrates were used because, in the initial study with catalyst 8, compounds with no terminal methyl groups (11, Scheme 4.3) or with trans methyl groups

(13, Scheme 4.3) underwent ARCM with low enantioselectivity. Many other substrates

74 were screened with the chiral ruthenium metathesis catalysts to determine if the substrate scope was general or limited, and the best results are shown in Table 4.5.

Table 4.5. Best results of ARCM reactions with “other” achiral trienes.

Triene Product Catalyst (mol %) ee (%) Conv. (%) O O 17a (2) 39 >98

16b (4) 64 9 11 12 Me Me Si Si O O 17a (2) 50 93

5 6

O 17a (2) 26 >98 O

43 16b (4) 61 9 46 Me Me Si a Si O 18a (2) –8 >98 O 16b (4) 15 18 47 45 Me Me Si Si O O >98 50 (5) N/A complex mixture

48 49 Me Me Si Si O O 50 (5) N/A 18%

51 52 O O O O >98% 50 (5) N/A complex mixture

53 54 O O 8b (4) 29 >98

55 56

Conditions: dichloride catalyst (2 mol %), triene, CH2Cl2 (0.055 M), 40 °C, 2 h; or dichloride catalyst (4 mol %), NaI (25 equiv relative to catalyst), triene, THF (0.055 M), 40 °C, 2 h. a See appendix 1 for information on an interesting side product. Catalyst 50 = (H2IMes)RuCl2(=CHPh). N/A = not applicable.

75 The meta substituted catalysts (16 and 17) were more selective than 8 in reactions with 11, but 12 was formed in either poor ee and excellent conversion or moderate ee and poor conversion. Compound 5, which is analogous to the excellent substrate 26, was only formed in 50% ee. Just as in the reactions with substrates 11 and 5, trienes 46 and

47 did not undergo ARCM as selectively and efficiently as the analogous compounds with cis methyl groups.15 The lower conversions obtained using diiodide catalysts with

11, 46, and 47 may be due to the formation of a ruthenium methylidene instead of an ethylidene that would be generated with 3, 34, and 35. Ruthenium methylidenes are known to decompose more rapidly than other alkylidenes, and in these cases decomposition may have occurred more quickly than the cross-metathesis reaction that introduced the triene to the catalyst.16

Substrates other than those lacking terminal methyl groups were also explored.

The RCM of 48 with the achiral catalyst 50 ((H2IMes)RuCl2(=CHPh)) afforded a mixture of volatile products, but no starting material remained; secondary metathesis with the acyclic olefin in 49 likely occurred. Vinylsilanes typically cause decomposition of ruthenium metathesis catalysts,17 so it was not surprising that very low yields of 52 were obtained when 51 was reacted with 50. Acrylate 53 reacted with catalyst 50, but a mixture of unidentified products were formed. Finally, because only trienes with cis methyl groups underwent ARCM in high enantioselectivities, substrate 55 was made to test if groups larger than methyl would work. The product (56) was formed in only 29% ee. It was clear from these data that chiral catalysts 8, 16, 17, and 18 were only highly selective with a small set of substrates.

76 The conditions used in the ARCM reactions discussed above were very similar to the optimized conditions used in the initial report,4 so various factors were reexamined to attempt to improve yields and enantiomeric excesses. A solvent screen was performed first (Table 4.6). Reactions performed in acetone, N,N-dimethylformamide (DMF), and acetonitrile afforded the product in <10% yield. Ethyl acetate and MTBE were better solvents, but the yields were still low. The best solvents were methylene chloride, benzene, and THF, where 39 was formed in >98%. The enantioselectivity was not

affected when THF was used in place of CH2Cl2, and the slight reduction in benzene suggests that solvent coordination to the catalyst is not as important as in the molybdenum systems.18

Table 4.6. Solvent screen.

Si Si Si O catalyst (2 mol %) O O

solvent, 40 °C, 2 h 26 26 39

Solvent Catalyst (mol %) 26 (%) 39 (%) ee (%) acetone 50 >90 <10 N/A DMF 50 >90 <10 N/A acetonitrile 50 >90 <10 N/A t-butyl methyl ether 50 77 23 N/A ethyl acetate 50 67 33 N/A

Et2O 50 9 91 N/A CH2Cl2 18a <2 >98 92 benzene 18a <2 >98 88 THF 18a <2 >98 92 N/A = not applicable.

Another variable that was adjusted was the reaction temperature; it was lowered to 0 °C, and the conversions and enantioselectivities of the ARCM of 26 were explored

(Table 4.7). When catalysts 8a and 8b were used, the enantioselectivities increased relative to the reactions at 40 °C. On the other hand, the enantioselectivities decreased

77 when ARCM reactions with 16b and 18a were done. The highest conversion of 26 at

0 °C was 40%, and in all cases the reactions were allowed to proceed for 24 h with 5 mol

% of catalyst. It was suspected that catalyst initiation (phosphine dissociation) was slow at 0 °C, so acid was added in an attempt to increase the rate of initiation.19 Unfortunately, no 39 was formed upon addition of either 1 equivalent of HCl in diethyl ether or 20 equivalents of benzoic acid relative to the catalyst. Although in some reactions a small increase in enantioselectivity was observed, this approach was not practical because the conversions were low.

Table 4.7. ARCM reactions at 0 °C.

Si Si Si O catalyst (5 mol %) O O

CH2Cl2 or THF, T, 24 h 26 26 39

Catalyst (mol %) Reaction at 0 °C Reaction at 40 °C ee (%) Conv. (%) ee (%) Conv. (%) 8a 88 39 83 >98 8b 90 40 86 68 16b 88 29 90 >98 18a 87 25 92 >98

The last variable that was explored was a combination of the iodide source and the solvent in reactions with diiodide catalysts. The original protocol called for sodium

4 iodide in THF, but reactions with the dichloride catalysts were done in CH2Cl2, so sodium iodide and tetrabutylammonium iodide were tested in THF and CH2Cl2 (Table

4.8). By far the most selective combination is sodium iodide in THF. When sodium

iodide in CH2Cl2 is used in place of THF, the enantioselectivity suffers drastically, although it is not as low as the reaction with just 8a and no sodium iodide (35% ee). This

could be due to the insolubility of sodium iodide in CH2Cl2. After 1 h in THF, the

78 sodium iodide/8a mixture dissolved completely (by visual inspection), but a white solid

remained in CH2Cl2. The enantioselectivity also decreased when the iodide source was changed to tetrabutylammonium iodide, but it remained higher than the reaction without

any iodide. The [Bu4N]I never completely dissolved in THF, but it appeared to be completely soluble in CH2Cl2. In both reactions 4 was formed in 59% ee. One explanation for the success of the NaI/THF combination is that NaI is soluble in THF but

NaCl is not; so the equilibrium between I-bound and Cl-bound ruthenium is forced to generate only the ruthenium diiodide catalyst.

Table 4.8. Effects of iodide source and solvent in ARCM.

O O 8a (4 mol %)

solvent, I- salt (25 equiv relative to 8a) 40 °C, 2 h 3 4

Iodide Salt Solvent ee (%) Conv. (%)

none CH2Cl2 35 >98 NaI THF 90 >98

NaI CH2Cl2 46 >98 [Bu4N]I THF 59 >98 [Bu4N]I CH2Cl2 59 >98

Absolute Stereochemistry Proof. All four chiral ruthenium catalysts tested in this study

(8, 16, 17, and 18) afforded the same enantiomer of the product for any given substrate in

Tables 4.1–4.4, and the absolute stereochemistry of a few of the products was determined

(Scheme 4.9). The absolute stereochemistry of 12-ent was proven by an independent synthesis using a Sharpless kinetic resolution.2 Compound 12-ent was also made with chiral molybdenum catalysts, and a GC trace using a Chiraldex GTA column was used to determine the enantioselectivities of the reactions. The chiral ruthenium complex 8a catalyzed the formation of 12, which was also separated on a Chiraldex GTA column.

79 The GC traces showed that chiral catalyst 8a, and therefore all the catalysts used in this study, afforded 12 in the absolute configuration shown in Scheme 4.9 and Table 4.5. The absolute configuration of 4 was determined by exposing it to ethenolysis conditions, which generated the same enantiomer of 12 that was obtained by the ARCM reaction of

11. Finally, compound 39 was oxidized to diol 57 and exposed to a one-pot mesylation/intramolecular nucleophilic displacement sequence to afford the same enantiomer of 4 that was obtained by the ARCM of 3. Chiral, cyclic products derived from a triene lacking terminal methyl groups (11), containing cis methyl groups and an alkenyl ether (3), and containing cis methyl groups and a dimethylsilyl alkenyl ether (26) all had the same absolute stereochemistry, which suggests that all of the products synthesized with 8, 16, 17, or 18 have the same absolute configurations.

O 8a 2 mol % O O

CH2Cl2, 40 °C, 2 h 11 12 12-ent 39% ee ref. 2 Chiral GTA column Same GC trace Ph Ph Chiral GTA column Me i-Pr N N MeO OMe i-Pr Cl Ru Me Cl Ph O 2 mol % O O PCy3 50 (5 mol %)

60 psi ethylene CH2Cl2, 40 °C, 2 h CH2Cl2, 40 °C 3 4 16 h 12 35% ee 33% ee Same GC trace Chiral GTA column

KF OH MsCl (1 equiv) Si Si KHCO3 O O 16a (3 mol %) O OH H2O2 Et3N

NaI, THF, 40 °C, 2 h MeOH/THF CH2Cl2 (1:1) rt, 15 min 26 39 rt, 16 h 57 4 87% ee 86% ee

Scheme 4.9. Absolute stereochemistry proof.

80 Model to Rationalize Enantioselectivity of Chiral Ruthenium Catalysts. An understanding of how and why catalysts 8, 16, 17, and 18 induce asymmetry in the RCM of achiral trienes would allow for the rational design of new, more selective catalysts.

Therefore, a model to explain the experimental observations was developed. First, it was suspected that the ARCM reactions were irreversible, so enantioenriched 39 was treated with achiral catalyst 50 under the same conditions used in the ARCM reactions. No erosion in the enantiomeric excess of 39 was detected. Additional support for the irreversibility of ARCM came from the fact that essentially no loss in enantiomeric excess was observed when 4 was reacted with ethylene and achiral catalyst 50 under forcing conditions (Scheme 4.10).

O O 50 (5 mol %) O

60 psi ethylene CH2Cl2 4 40 °C, 16 h 4 12 35% ee 35% ee 33% ee 83% unreacted 17% conv.

Scheme 4.10. Ethenolysis of 4.

Chapters 2 and 3 of this dissertation describe how the achiral variant (50) of the chiral catalysts used here does not readily react with 1,1-disubstituted and trisubstituted alkenes. Therefore, all of the ARCM reactions are thought to proceed through a ruthenium alkylidene derived from the monosubstituted olefin present in every substrate.

This species binds one of the diastereotopic alkenes and, through a metallacyclobutane intermediate/transition state, forms the ring-closed product.

Olefin coordination to the ruthenium is an important, stereodefining step; unfortunately, the actual position where the olefin binds relative to the NHC is unclear.

There is experimental evidence to support coordination both cis and trans to the NHC.20

If coordination is cis, than the alkene should bind to the catalyst face opposite the

81 isopropyl group to avoid an interaction with it (Scheme 4.11, 61).4 Either olefin could coordinate cis to the NHC, but structure 61 has the non-binding alkene in the more stable, pseudoequatorial position of the cyclic intermediate. Additionally, a hydrogen instead of a vinyl group is directed toward a halide. Completion of the ARCM reaction from 61 affords the major enantiomer.

trans olefin binding pathway

Ph Ph Ph Ph Ph Ph i-Pr i-Pr i-Pr N N N N N N i-Pr X i-Pr X i-Pr X Ru Ru O Ru O X X O X

58 59

O O minor enantiomer major enantiomer

4-ent 4

Ph Ph Ph Ph Ph Ph i-Pr i-Pr i-Pr N N N N N N X i-Pr X i-Pr i-Pr Ru Ru Ru O X O O X X X H 60 61 cis olefin binding pathway

Scheme 4.11. Proposed pathways leading to the desired product.

Recent experimental reports suggest the alkene may be coordinating trans to the

NHC,20b,c and computational studies also support trans binding.21 One computational study explored the energetics of 8b reacting with 3, and the most favored pathway is depicted in Scheme 4.11 (trans olefin binding pathway).21a The steric interaction that favors 59 over 58 is between the substituents on the alkylidene and the N-bound aryl ring.

82 The calculations concluded that the aryl ring is not orthogonal to the plane of the NHC

(90°), but is instead tilted 75°. X-ray crystallographic analysis of a derivative of 7a provided experimental evidence to support the calculations.4 In this model, the side of the aryl ring with an isopropyl group was found to be smaller than the side with no substitution, so the larger alkylidene substituent (the ether) was positioned under the isopropyl group (Figure 4.2). Just as in 61, the non-coordinating vinyl group is in the pseudoequatorial position in 59.

Ph Ph Ph i-Pr i-Pr N N H X X i-Pr X H Ru O O X

alkylidene under i-Pr group

Figure 4.2. Suggested alkylidene position in trans olefin binding pathway.

Although the discussion presented here focuses on a substrate that forms a five- membered ring, the ideas could be extended to trienes that form other ring sizes. For a given substrate, there is presumably an energetically favored ring conformation that occurs once the olefin is coordinated to the ruthenium center. By adding substitution to the ring, the difference in energies of the cyclic, olefin-bound intermediates/transition states may increase, and the reaction may be more selective. That is a possible explanation as to why the reactions with the dimethylsilyl-containing trienes were more enantioselective than the reactions with substrates containing only alkyl ethers.

In almost every reaction shown in Tables 4.1–4.3, the addition of sodium iodide increased the enantioselectivity. One explanation for this effect is based on steric interactions: in the cis olefin binding pathway of Scheme 4.11, either a hydrogen (as

83 shown in 61) or a vinyl group could be directed at the halide trans to the NHC. If the vinyl group was directed at the halide, the minor enantiomer would be formed. As the size of the halide is increased from chloride to iodide, the strength of the interaction between the two groups is be expected to increase, and the structure with a hydrogen in that position is expected to be lower in energy. In the trans olefin binding pathway, calculations suggest that as the halides increase in size, they are pushed away from each other and toward the alkylidene. That would put them in closer proximity to the reacting olefin, creating a smaller binding pocket, and therefore a more selective reaction.21a

Another explanation for the large iodide effect is based on electronic factors. It is known that phosphine dissociation occurs more quickly for the diiodide variant of achiral catalyst 50 than for 50-dichloride, but the reactivities of the active species are similar.13b

Phosphine dissociation is the rate-limiting step in catalyst initiation, so the fact that 50- diiodide initiates quickly but does not increase the rate of product formation suggests that the active species derived from 50-diiodide is less active than that derived from 50- dichloride. Based on these data, it is reasonable to assume that the active species of the chiral diiodide catalysts are less active than the dichloride active species. When a highly active catalyst with an alkylidene derived from one of the achiral trienes discussed above is used, the rate- and enantio-determining step might be olefin binding. Therefore, metathesis would occur regardless of which diastereotopic alkene initially coordinated to the ruthenium center. On the other hand, a less-active catalyst may coordinate the olefin but not perform the ring-closing reaction. Instead, the olefin could dissociate and re-bind, setting up a rapid equilibrium. In this case ruthenacyclobutane formation, not olefin binding, would be the rate- and enantio-determining step, regardless of which conformer

84 is more stable. A change in the rate-determining step would also change the ΔΔG‡, thereby affording a different distribution of enantiomeric products relative to the reactions catalyzed by a more active alkylidene.22,23

Two models have been proposed to explain the origin of enantioselectivity in reactions with the chiral ruthenium catalysts used in this study: one assuming the olefin coordinates cis to the NHC, and one assuming it coordinates trans to the NHC. At this point, neither of these pathways can be ruled out because no experiments that provide solid insight into catalyst structure during turnover have been developed. Both a steric and an electronic justification have been presented to explain why the diiodide catalysts are more selective than the dichloride catalysts. Unfortunately, no model has been developed that clarifies why the substrates containing trisubstituted alkenes with cis methyl groups afford products with much higher enantiomeric excesses.

Conclusion

Using novel, asymmetric, ruthenium metathesis catalysts containing chiral, monodentate NHCs, achiral trienes were desymmetrized in up to 92% ee. Catalysts 16 and 17, which contained substitution para to an ortho isopropyl group, behaved very similarly to the parent chiral catalyst 8. When the chloride ligands were exchanged for iodides, the enantioselectivities increased drastically in many reactions with all three catalysts. Complex 18 was the most selective catalyst, and it could be used in loadings of

1% or less to afford the desired products in up to 92% ee and in high conversions. Many achiral trienes with varying substitution were explored, and only those with two trisubstituted olefins with cis methyl groups afforded the desired products with high

85 enantioselectivities. Two proposed models that account for the observed products were presented, as well as two explanations as to why the diiodide catalysts are more selective than the dichloride catalysts. Although the substrate scope is limited and further development is needed to make these catalysts more general, the reactions are procedurally simple and very reliable due to the air and moisture stability of the ruthenium alkylidene catalysts.

Experimental

General Information. NMR spectra were recorded on an Oxford 300 MHz NMR spectrometer running Varian VNMR software. Chemical shifts are reported in parts per million (ppm) downfield from tetramethylsilane (TMS) with reference to internal solvent for 1H NMR and 13C NMR spectra. Multiplicities are abbreviated as follows: singlet (s), doublet (d), triplet (t), quartet (q), quintet (quint), sextet (sext), septet (sept), multiplet

(m), and broad (br). Optical rotations were taken on a Jasco P-1010 polarimeter with a wavelength of 589 nm. The concentration “c” has units of g/100 mL (or 10 mg/mL) unless otherwise noted. Analytical thin-layer chromatography (TLC) was performed using silica gel 60 F254 precoated plates (0.25 mm thickness) with a fluorescent indicator. Visualization was performed with standard potassium permanganate stain (10

g KMnO4, 20 g Na2CO3, 1 L water) or UV light. Flash column chromatography was performed using silica gel 60 (230–400 mesh). All enantiomeric purities were determined by chiral GC (Chiraldex G-TA, 30 m × 0.25 mm or CP Chirasil-Dex-CB, 25

m × 0.25 mm) or chiral SFC (supercritical CO2, ADH column, 214 nm UV detection) and were compared to racemic samples. All glassware was either oven dried or flame dried,

86 and reactions were done under an atmosphere of argon unless otherwise noted. All organic solvents were dried by passage through solvent purification columns containing activated alumina. All commercial chemicals were used as obtained. Compounds 11,3

12,3 5,3 6,3 46,24 43,24 47,24 and 4524 are known compounds.

(2E,5E)-3,5-Dimethylhepta-2,5-dien-4-ol (21).11 Titanocene dichloride (444 mg, 1.78 mmol) was added to a solution of 2-butyne (22) (5.6 mL, 3.9 g, 71 mmol) and

isobutylmagnesium bromide (2.0 M in diethyl ether, 33 mL, 66 mmol) in 60 mL Et2O, and the solution stirred at rt for 1 h. Trans-2-methyl-2-butenal (20) (5.7 mL, 5.0 g, 59

mmol) in 30 mL Et2O was added slowly, and the mixture stirred at rt for 3 h. It was

quenched with saturated aqueous NH4Cl (100 mL), filtered through a pad of Celite, and the organic layer was removed from the filtrate. The aqueous layer was extracted with ether (3 × 75 mL), and the organic layers were combined, washed with brine, dried over

MgSO4, and evaporated to a brown oil. The oil was purified by flash chromatography

(10% EtOAc in hexanes) to a yellow oil, which was distilled (Kugelrohr, 1 torr, 120 °C)

1 to give 7.20 g (86% yield) of 21 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ

5.56 (qquint, J = 6.6, 1.4 Hz, 2H), 4.34 (s, 1H), 1.63 (dt, J = 6.9, 1.1 Hz, 6H), 1.47 (t, J =

13 1.1 Hz, 6H). C NMR (75 MHz, CDCl3, ppm): δ 136.1, 120.4, 81.8, 13.3, 12.1. HRMS

(EI) m/z calc. for C9H16O: 140.1201, found 140.1203.

(2E,5E)-4-(Allyloxy)-3,5-dimethylhepta-2,5-diene (3). Alcohol 21 (200 mg, 1.43 mmol) was added dropwise to a suspension of NaH (60% in oil, 114 mg, 2.85 mmol) in

6 mL THF. After stirring at reflux for 15 min, the mixture was allowed to cool to rt, and

87 allyl bromide (430 mg, 3.57 mmol) was added. The mixture stirred at reflux for 4 h, was

quenched with saturated aqueous NH4Cl (10 mL), and was extracted with ether (3 × 15

mL). The organic layers were combined, dried over MgSO4, and evaporated to an oil which was purified by flash chromatography (1% EtOAc in hexanes) to give 210 mg

(82% yield) of 3 as a colorless oil. Spectral data matched those in the literature.2 1H

NMR (300 MHz, CDCl3, ppm): δ 5.85–5.98 (m, 1H), 5.55 (qq, J = 6.6, 1.1 Hz, 2H),

5.22–5.29 (m, 1H), 5.10–5.15 (m, 1H), 3.94 (br s, 1H), 3.85 (dq, J = 5.5, 0.8 Hz, 2H),

13 1.63 (dq, J = 6.6, 1.1 Hz, 6H), 1.46 (d, J = 1.1 Hz, 6H). C NMR (75 MHz, CDCl3, ppm): δ 135.6, 134.2, 121.2, 116.3, 88.3, 68.8, 13.3, 12.3. HRMS (EI) m/z calc. for

C12H20O: 180.15142, found 180.15135.

(2E,5E)-3,5-Dimethyl-4-(pent-4-enyloxy)hepta-2,5-diene (23). 21 (400 mg, 2.9 mmol) was added slowly to a suspension of NaH (60% in oil, 140 mg, 3.4 mmol) in 5 mL THF at rt, and some bubbling occurred. After 2.5 h at rt, 5-bromo-1-pentene (0.68 mL, 5.7 mmol) was added, and the mixture was heated to reflux for 16 h. It was cooled to rt, carefully quenched with 20 mL water, and extracted with ether (3 × 25 mL). The organic

layers were combined, washed with brine, dried over Na2SO4, and evaporated to an oil, which was purified by flash chromatography (1% EtOAc in hexanes, then 10% EtOAc in hexanes) to give 130 mg (22% yield) of 23 as a colorless oil. 1H NMR (300 MHz,

CDCl3, ppm): δ 5.76–5.90 (m, 1H), 5.53 (qquint, J = 6.6, 1.4 Hz, 2H), 4.92–5.04 (m, 2H),

3.86 (s, 1H), 3.29 (t, J = 6.6 Hz, 2H), 2.13 (q, J = 6.9 Hz, 2H), 1.63–1.71 (m, 2H), 1.63

13 (d, J = 6.6 Hz, 6H), 1.46 (s, 6H). C NMR (75 MHz, CDCl3, ppm): δ 138.9, 134.6,

120.9, 114.6, 89.1, 67.5, 30.8, 29.4, 13.4, 12.3. HRMS (EI) m/z calc. for C14H24O:

208.1827, found 208.1828.

(2E,5E)-4-(Hex-5-enyloxy)-3,5-dimethylhepta-2,5-diene (24). 21 (500 mg, 3.6 mmol) was added slowly to a suspension of NaH (60% in oil, 285 mg, 7.1 mmol) in 7 mL THF at rt, and some bubbling occurred. After 15 min at rt, 6-bromo-1-hexene (0.96 mL, 7.1 mmol) was added, the mixture was heated to reflux for 16 h. It was cooled to rt, carefully

quenched with 20 mL saturated aqueous NH4Cl, and extracted with ether (3 × 25 mL).

The organic layers were combined, washed with brine, dried over MgSO4, and evaporated to an oil, which was purified by flash chromatography (2% EtOAc in hexanes) to give 346 mg (44% yield) of 24 as a colorless oil. 1H NMR (300 MHz,

CDCl3, ppm): δ 5.74–5.88 (m, 1H), 5.53 (qt, J = 6.6, 1.1 Hz, 2H), 4.91–5.03 (m, 2H),

3.85 (s, 1H), 3.27 (t, J = 6.4 Hz, 2H), 2.06 (q, J = 7.1 Hz, 2H), 1.63 (d, J = 6.9 Hz, 6H),

13 1.54–1.60 (m, 2H), 1.40–1.52 (m, 2H), 1.45 (s, 6H). C NMR (75 MHz, CDCl3, ppm): δ

139.2, 134.6, 120.9, 114.5, 89.1, 67.9, 33.9, 29.6, 25.9, 13.3, 12.3. HRMS (EI) m/z calc.

for C15H26O: 222.1984, found 222.1971.

Allyl(((2E,5E)-3,5-dimethylhepta-2,5-dien-4-yloxy)methyl)dimethylsilane (25). 21

(300 mg, 2.1 mmol) was added to a suspension of NaH (60% in oil, 103 mg, 2.6 mmol) in 3 mL THF and some bubbling occurred. After 30 min at rt, allylchloromethyldimethylsilane (0.70 mL, 0.63 g, 4.3 mmol) was added, and the mixture was heated to reflux for 16 h. The reaction mixture was cooled to rt, quenched with 20 mL water, and extracted with ether (3 × 20 mL). The organic layers were combined,

89 washed with brine, dried over MgSO4, and evaporated to a yellow oil, which was purified by flash chromatography (100% hexanes) to give 349 mg (65% yield) of 25 as a colorless

1 oil. H NMR (300 MHz, CDCl3, ppm): δ 5.81–5.95 (m, 1H), 5.51 (qquint, J = 6.6, 1.4

Hz, 2H), 4.98 (dq, J = 17.1, 1.7 Hz, 1H), 4.85–4.91 (m, 1H), 4.28 (s, 1H), 2.03–2.12 (m,

2H), 1.61 (dt, J = 6.9, 1.1 Hz, 6H), 1.43 (t, J = 1.1 Hz, 6H), 0.64–0.70 (m, 2H), 0.07 (s,

13 6H). C NMR (75 MHz, CDCl3, ppm): δ 142.0, 136.5, 119.8, 112.8, 82.3, 27.6, 16.2,

13.3, 12.1, –1.4. HRMS (FAB) m/z calc. for C15H28OSi: 252.1910, found 252.1914.

Allyl((2E,5E)-3,5-dimethylhepta-2,5-dien-4-yloxy)dimethylsilane (26).

Allylchlorodimethylsilane (1.1 mL, 7.5 mmol) was added to a solution of 21 (1.0 g, 7.1 mmol), triethylamine (1.2 mL, 8.6 mmol), and N,N-dimethylaminopyridine (44 mg, 0.4

mmol) in 30 mL CH2Cl2 at rt. After 5 h the reaction was quenched with 50 mL water, the organic layer was removed, and the aqueous layer was extracted with ether (3 × 50 mL).

The organic layers were combined, washed with brine, dried over Na2SO4, and evaporated to an oil. The oil was redissolved in hexanes and was filtered through a pad of neutral alumina. The filtrate was condensed to give 1.30 g (76% yield) of 26 as a colorless oil. Attempts to purify 26 by silica gel chromatography resulted in inconsistent yields and varying levels of purity due to product decomposition. 1H NMR (300 MHz,

CDCl3, ppm): δ 5.70–5.85 (m, 1H), 5.52 (qquint, J = 6.9, 1.4 Hz, 2H), 4.80–4.90 (m, 2H),

4.30 (s, 1H), 1.61 (dt, J = 6.9, 1.1 Hz, 6H), 1.58–1.63 (m, 2H), 1.43 (t, J = 1.1 Hz, 6H),

13 0.08 (s, 6H). C NMR (75 MHz, CDCl3, ppm): δ 136.4, 134.8, 119.9, 113.5, 82.4, 25.1,

13.3, 12.0, –1.9. HRMS (EI) m/z calc. for C14H26OSi: 238.1753, found 238.1752.

MgBr Ph Ph Ph Ph Si Si H Cl THF, 0 °C to 40 °C H 62 75% yield 63

Allyldiphenylsilane (63).25 To a solution of 62 (2.0 g, 1.8 mL, 9.1 mmol) in 9 mL of

THF at 0 °C was added allylmagnesium bromide (1.0 M in diethyl ether, 9.7 mL, 9.7 mmol) over 5 min. After 10 min at 0 °C, the cloudy mixture was warmed to 40 °C. After

2 h at 40 °C, the reaction was quenched with a few pieces of ice followed by 30 mL saturated aqueous ammonium chloride. It was extracted with diethyl ether (3 × 30 mL),

dried over MgSO4, and concentrated to an oil. Purification by flash chromatography

(100% pentane) afforded 1.53 g (75% yield) of 63 as a colorless oil. 1H NMR (300 MHz,

CDCl3, ppm): δ 7.56–7.60 (m, 4H), 7.34–7.45 (m, 6H), 5.79–5.93 (m, 1H), 4.96 (dq, J =

17.1, 1.7 Hz, 1H), 4.88–4.93 (m, 1H), 4.87 (t, J = 3.6 Hz, 1H), 2.15 (dd, J = 7.7, 3.6 Hz,

2H).

CuCl2 Ph Ph CuI (2.5 mol %) Ph Ph Si Si H THF, rt Cl 63 45% yield 64

Allylchlorodiphenylsilane (64).26 A two-neck round-bottom flask topped with a

Schlenk filter connected to another round-bottom flask was charged with anhydrous

CuCl2 (1.1 g, 8.0 mmol) and anhydrous CuI (19 mg, 0.10 mmol). After 2 pump/backfills with argon, the powders were suspended in 8 mL of THF, allyldiphenylsilane (63) (0.90 g, 4.0 mmol) was added, and the orange/brown slurry was stirred vigorously. After 16 h at rt, the reaction mixture was completely colorless and a white suspension was present.

The round-bottom/Schlenk filter apparatus was inverted and placed under slight vacuum, and the filtrate was concentrated to a viscous oil. Purification by distillation (Kugelrohr,

0.4 torr, 210 °C) afforded 0.47 g (45% yield) of 64 as a pale yellow oil. 1H NMR (300

MHz, CDCl3, ppm): δ 7.63–7.67 (m, 4H), 7.38–7.50 (m, 6H), 5.75–5.90 (m, 1H), 5.00–

5.05 (m, 1H), 4.97–5.00 (m, 1H), 2.36 (dt, J = 8.0, 1.4 Hz, 2H).

Allyl((2E,5E)-3,5-dimethylhepta-2,5-dien-4-yloxy)diphenylsilane (27). To a solution of allylchlorodiphenylsilane (64) (370 mg, 1.43 mmol), triethylamine (0.28 mL, 2.0

mmol), and N,N-dimethylaminopyridine (8.7 mg, 0.07 mmol) in 7 mL CH2Cl2 was added

21 (241 mg, 1.72 mmol). After 5 h at rt, the reaction was poured into 20 mL of water, the

organic layer was removed, and the aqueous layer was extracted with CH2Cl2 (3 × 20

mL). The organic layers were combined, washed with brine, dried over MgSO4, and evaporated to an oil. Purification by flash chromatography afforded 359 mg (69% yield)

1 of 27 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 7.55–7.59 (m, 4H), 7.31–

7.39 (m, 6H), 5.71–5.85 (m, 1H), 5.50 (tq, J = 6.9, 1.4 Hz, 2H), 4.81–4.91 (m, 2H), 4.39

(br s, 1H), 2.15 (dt, J = 8.0, 1.4 Hz, 2H), 1.56 (dt, J = 6.9, 0.8 Hz, 6H), 1.39 (t, J = 1.1 Hz,

13 6H). C NMR (75 MHz, CDCl3, ppm): δ 135.89, 135.32, 135.11, 133.62, 129.86,

127.74, 120.29, 114.97, 83.04, 22.48, 13.31, 11.94. HRMS (EI) m/z calc. for C24H30SiO

[M+] 362.2066, found 362.2077.

(2E,5E)-4-(But-3-enyloxy)-3,5-dimethylhepta-2,5-diene (29). 3 (1.07 g, 5.90 mmol) in

3.6 mL THF was added to a solution of 9-BBN (0.5 M in THF, 14.2 mL, 7.12 mmol), and the solution stirred at rt. After 5 h 3.6 mL ethanol was added, followed by 1.4 mL

aqueous 6 M NaOH and 2.8 mL 30% H2O2, and the reaction stirred at 50 °C for 1 h. It was diluted with 20 mL saturated aqueous NaHCO3 and was extracted with ether (3 × 25

mL). The organic layers were combined, washed with brine, dried over Na2SO4,

92 evaporated to an oil, and purified by flash chromatography (20% EtOAc in hexanes) to

1 give 916 mg (83% yield) of 28 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ

5.51 (qquint, J = 6.9 Hz, 1.1 Hz, 2H), 3.89 (br s, 1H), 3.79 (t, J = 5.2 Hz, 2H), 3.50 (t, J =

5.5 Hz, 2H), 2.62 (br s, 1H), 1.84 (quint, J = 5.8 Hz, 2H), 1.64 (dt, J = 6.6 Hz, 1.1 Hz,

6H), 1.46 (t, J = 1.4 Hz, 6H). DMSO (0.89 mL, 12.6 mmol) was added slowly to a

solution of oxalyl chloride (0.66 mL, 7.56 mmol) in 15 mL CH2Cl2 at –78 °C. After 5 min a solution of 28 (500 mg, 2.52 mmol) in 5 mL CH2Cl2 was added to the –78 °C reaction solution, and it stirred for 30 min. Triethylamine (2.5 mL, 17.6 mmol) was added, and after 30 min at –78 °C, the reaction slowly warmed to rt. It was quenched with 40 mL water and extracted with ether (3 × 50 mL). The organic layers were

combined, dried over Na2SO4, and evaporated to 448 mg of the crude aldehyde as an orange oil, which was used in the next step without further purification (attempts to purify this aldehyde by silica gel chromatography resulted in product decomposition and low (~30%) isolated yields). To a suspension of triphenylmethylphosphonium bromide

(2.15 g, 6.0 mmol) in 20 mL THF at 0 °C was added n-butyllithium (2.5 M in hexanes,

2.0 mL, 5.0 mmol). After 20 min a solution of the crude aldehyde (448 mg, 2.3 mmol) in

5 mL THF was added slowly to the orange reaction mixture, and it stirred at 0 °C for 1 h.

It was quenched with 30 mL saturated aqueous NH4Cl and extracted with ether (3 × 25

mL). The organic layers were combined, dried over MgSO4, and evaporated to an oil, which was purified by flash chromatography (1% EtOAc in hexanes) to give 179 mg of a

colorless oil. To a solution of the oil in 10 mL CH2Cl2 was added 3% hydrogen peroxide, and the mixture was shaken for 15 minutes. The organic layer was removed, dried over

Na2SO4, evaporated to an oil, and filtered through a plug of silica gel (1% EtOAc in

93 hexanes). The filtrate was concentrated to 145 mg (29% yield over 2 steps) of 29 as a

1 colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 5.77–5.92 (m, 1H), 5.53 (qquint, J =

6.9, 1.4 Hz, 2H), 4.98–5.11 (m, 2H), 3.88 (s, 1H), 3.33 (t, J = 6.9 Hz, 2H), 2.33 (q, J = 6.9

13 Hz, 2H), 1.63 (d, J = 6.6 Hz, 6H), 1.46 (s, 6H). C NMR (75 MHz, CDCl3, ppm): δ

136.0, 134.5, 121.0, 116.2, 89.1, 67.6, 34.7, 13.3, 12.3. HRMS (EI) m/z calc. for

C13H22O: 194.1671, found 194.1679.

(2E,7E)-3,7-Dimethylnona-2,7-dien-5-ol (33). Titanocene dichloride (212 mg, 0.85 mmol) was added to a solution of 2-butyne (2.4 mL, 30 mmol) and isobutylmagnesium

bromide (2.0 M in diethyl ether, 15 mL, 30 mmol) in 30 mL Et2O, and the solution stirred at rt for 1 h. This brown solution was slowly transferred via syringe to a suspension of

CuBr (397 mg, 2.8 mmol) in Et2O (75 mL) at –78 °C. After 5 min epichlorohydrin (30)

(2.2 mL, 28 mmol) was added slowly to the mixture. It stirred at –78 °C for 3 h, and was allowed to warm to –40 °C where it continued stirring for 48 h. The reaction mixture was poured into 100 mL aqueous 1 N HCl and was extracted with ether (3 × 100 mL).

The organic layers were combined, dried over MgSO4, and evaporated to an oil, which was purified by flash chromatography (10% EtOAc in hexanes) to give 2.83 g (68%

1 yield) of the chlorohydrin 31 as a yellow oil. H NMR (300 MHz, CDCl3, ppm): δ 5.36

(qq, J = 6.6, 1.1 Hz, 1H), 3.89–3.97 (m, 1H), 3.61 (dd, J = 11.0, 4.1 Hz , 1H), 3.50 (dd, J

= 11.0, 6.3 Hz, 1H), 2.29 (dd, J = 13.5, 5.5 Hz, 1H), 2.20 (dd, J = 13.5, 8.0 Hz, 1H), 2.04

(br s, 1H), 1.66 (t, J = 1.1 Hz, 3H), 1.62 (dt, J = 6.9 Hz, 0.8 Hz, 3H). The chlorohydrin

31 (2.8 g, 19 mmol) was added slowly to a suspension of NaH (60% in oil, 1.13 g, 28 mmol) in 50 mL THF, and the mixture stirred at reflux for 16 h. It was cooled to rt,

94 quenched with saturated aqueous NH4Cl until pH = 9 was reached, and was extracted with ether (3 × 50 mL). The organic layers were combined, dried over MgSO4, and

evaporated to a yellow oil, which was purified by flash chromatography (1% Et2O in pentane) to give 1.08 g (51% yield) of the epoxide 32 as a pale yellow oil. 1H NMR (300

MHz, CDCl3, ppm): δ 5.34 (qq, J = 6.6, 1.1 Hz, 1H), 2.96–3.02 (m, 1H), 2.77 (dd, J =

4.9, 3.8 Hz, 1H), 2.49 (dd, J = 4.9, 2.7 Hz, 1H), 2.25 (dd, J = 14.8, 6.0 Hz, 1H), 2.16 (dd,

J = 14.5, 5.5 Hz, 1H), 1.69 (t, J = 1.1 Hz, 3H), 1.61 (dq, J = 6.6, 1.1 Hz, 3H). Titanocene dichloride (69 mg, 0.28 mmol) was added to a solution of 2-butyne (0.8 mL, 10 mmol) and isobutylmagnesium bromide (2.0 M in diethyl ether, 4.9 mL, 10 mmol) in 10 mL

Et2O, and the solution stirred at rt for 1 h. This brown solution was slowly transferred via

syringe to a suspension of CuBr (128 mg, 0.9 mmol) in Et2O (25 mL) at –78 °C. After 5 min the epoxide 32 (1.0 g, 9 mmol) was added slowly to the mixture. It stirred at –78 °C for 2 h, and was allowed to warm to –40 °C where it continued stirring for 24 h. The reaction mixture was poured into 75 mL aqueous 1 N HCl and was extracted with ether

(3 × 50 mL). The organic layers were combined, dried over Na2SO4, and evaporated to an oil, which was purified by flash chromatography (7% EtOAc in hexanes) to give 894 mg (60% yield, 21% over 3 steps) of 33 as a pale yellow oil. 1H NMR (300 MHz,

CDCl3, ppm): δ 5.30 (q, J = 6.6 Hz, 2H), 3.75–3.82 (m, 1H), 2.01–2.15 (m, 4H), 1.79 (d,

13 J = 1.7 Hz, 1H), 1.62 (s, 6H), 1.59 (d, J = 6.6 Hz, 6H). C NMR (75 MHz, CDCl3, ppm):

δ 133.0, 122.2, 66.6, 47.7, 16.0, 13.7. HRMS (EI) m/z calc. for C11H20O: 168.1514, found

168.1515.

95 (2E,7E)-5-(Allyloxy)-3,7-dimethylnona-2,7-diene (34). Alcohol 33 (200 mg, 1.2 mmol) was added dropwise to a suspension of NaH (60% in oil, 95 mg, 2.4 mmol) in 5 mL THF.

After stirring at reflux for 15 min, the mixture was allowed to cool to rt, and allyl bromide (360 mg, 3.0 mmol) was added. The mixture stirred at reflux for 12 h, was

quenched with saturated aqueous NH4Cl (30 mL), and was extracted with ether (3 × 30

mL). The organic layers were combined, dried over MgSO4, and evaporated to an oil which was purified by flash chromatography (1.5% EtOAc in hexanes) to give 180 mg

1 (73% yield) of 34 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 5.80–5.93 (m,

1H), 5.20–5.28 (m, 3H), 5.09–5.13 (m, 1H), 3.98 (dt, J = 5.4, 1.5 Hz, 2H), 3.50–3.58 (m,

1H), 2.18 (dd, J = 13.2, 6.9 Hz, 2H), 2.07 (dd, J = 13.5, 6.0 Hz, 2H), 1.62 (t, J = 1.2 Hz,

6H), 1.56–1.59 (m, 6H).

Allyl((2E,7E)-3,7-dimethylnona-2,7-dien-5-yloxy)dimethylsilane (35). To a solution of 33 (150 mg, 0.9 mmol), triethylamine (0.25 mL, 1.8 mmol), and N,N-

dimethylaminopyridine (5 mg, 0.04 mmol) in 5 mL CH2Cl2 was added allylchlorodimethylsilane (0.20 mL, 1.3 mmol). After stirring at rt for 16 h, the reaction was quenched with 10 mL water and extracted with diethyl ether (3 × 20 mL). The

organic layers were combined, washed with brine, dried over Na2SO4, and evaporated to an oil, which was purified by flash chromatography (1% EtOAc in hexanes) to give 209

1 mg (88% yield) of 35 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 5.70–5.84

(m, 1H), 5.23 (qq, J = 6.6, 1.4 Hz, 2H), 4.81–4.89 (m, 2H), 3.85 (quint, J = 6.3 Hz, 1H),

2.05–2.08 (m, 4H), 1.60 (t, J = 1.1 Hz, 6H), 1.58 (dt, J = 6.6, 0.8 Hz, 6H), 1.54–1.57 (m,

13 2H), 0.05 (s, 6H). C NMR (75 MHz, CDCl3, ppm): δ 134.7, 133.0, 121.7, 113.5, 70.5,

48.2, 25.4, 16.5, 13.6, –1.7. HRMS (EI) m/z calc. for C16H30OSi: 266.2066, found

266.2070.

General Procedure A: Asymmetric Ring-Closing Reactions with 8a, 16a, 17a, and

18a. Triene was added to a solution of dichloride catalyst (1–2 mol %) in CH2Cl2 (0.055

M), and the reaction stirred at 40 °C for 2 h. The solvent was evaporated, and the remaining residue was purified by flash chromatography to yield the desired cyclic diene.

General Procedure B: Asymmetric Ring-Closing Reactions with 8b, 16b, 17b, and

18b. A solution of NaI (25 equiv. relative to catalyst) and dichloride catalyst (4 mol %) in THF was stirred at rt for 1 h. Triene (0.055 M) was added, and the solution stirred at

40 °C for 2 h. The solvent was evaporated, and the remaining residue was purified by flash chromatography to yield the desired cyclic diene.

(S,E)-2-(But-2-en-2-yl)-3-methyl-2,5-dihydrofuran (4). Following general procedure

B, 3 (40 mg, 0.22 mmol), 8a (8.9 mg, 0.0089 mmol), and NaI (33 mg, 0.22 mmol) in 4

mL THF gave 19.8 mg (64% yield) of 4 as a pale yellow oil (5% Et2O in pentane) in 90% ee. Chiraldex G-TA, 1mL/min, 60 °C for 60 min, retention times = 21.9 (major) and 23.4

25 1 (minor) min. [α]D = +116.5 (CHCl3, c = 0.55). H NMR (300 MHz, CDCl3, ppm): δ

5.56 (quint, J = 1.6 Hz, 1H), 5.52 (q, J = 6.9 Hz, 1H), 4.88 (br s, 1H), 4.53–4.68 (m, 2H),

1.64 (dq, J = 6.9, 1.1 Hz, 3H), 1.56 (quint, J = 1.4 Hz, 3H), 1.47 (quint, J = 1.4 Hz, 3H).

13 C NMR (75 MHz, CDCl3, ppm): δ 137.2, 135.6, 123.8, 121.5, 95.0, 75.6, 13.5, 12.4,

10.1. HRMS (EI) m/z calc. for C9H14O: 138.1045, found 138.1040.

(S,E)-6-(But-2-en-2-yl)-5-methyl-3,6-dihydro-2H-pyran (36). Following general procedure B, 29 (40 mg, 0.21 mmol), 8a (8.2 mg, 0.0082 mmol), and NaI (31 mg,

0.21 mmol) in 3.8 mL THF gave 24.1 mg (77% yield) of 36 as a pale yellow oil (3%

Et2O in pentane) in 90% ee. Chiraldex G-TA, 1mL/min, 60 °C for 60 min, retention

26 1 times = 30.6 (major) and 34.7 (minor) min. [α]D = +43.0 (CHCl3, c = 0.69). H NMR

(300 MHz, CDCl3, ppm): δ 5.64–5.68 (m, 1H), 5.53 (q, J = 6.6 Hz, 1H), 4.29 (s, 1H),

3.88–3.94 (m, 1H), 3.53–3.61 (m, 1H), 2.19–2.32 (m, 1H), 1.85–1.96 (m, 1H), 1.64 (dd, J

= 6.6, 1.1 Hz, 3H), 1.54 (t, J = 1.4 Hz, 3H), 1.47 (q, J = 1.4 Hz, 3H). 13C NMR (75 MHz,

CDCl3, ppm): δ 134.9, 134.5, 125.2, 121.5, 84.1, 62.9, 25.8, 19.7, 13.5, 11.5. HRMS (EI)

m/z calc. for C10H16O: 152.1201, found 152.1204.

(S,Z)-7-((E)-But-2-en-2-yl)-6-methyl-2,3,4,7-tetrahydrooxepine (37). Following a modified version of general procedure A, 23 (40 mg, 0.19 mmol) was added to a solution

of 18a (2.1 mg, 0.0019 mmol) in 3.5 mL CH2Cl2, and the reaction stirred at 40 °C. After

2 h, an additional portion of 18a (2.1 mg, 0.0019 mmol) was added, and the solution stirred at 40 °C for an additional 2 h. The solvent was removed by evaporation, and the residue was purified by flash chromatography (4% EtOAc in hexanes) to give 29.4 mg

(92% yield) of 37 as a yellow oil in 76% ee. Chiraldex G-TA, 1mL/min, 60 °C for 90

24 min, retention times = 75.1 (minor) and 76.6 (major) min. [α]D = +164.0 (CHCl3, c =

1 0.90). H NMR (300 MHz, CDCl3, ppm): δ 5.57–5.62 (m, 1H), 5.51 (q, J = 6.6 Hz, 1H),

4.26 (s, 1H), 3.85–3.92 (m, 1H), 3.57–3.66 (m, 1H), 2.48–2.59 (m, 1H), 1.87–2.03 (m,

2H), 1.67–1.80 (m, 1H), 1.67 (t, J = 2.2 Hz, 3H), 1.65 (d, J = 6.6 Hz, 3H), 1.50 (s, 3H).

98 13 C NMR (75 MHz, CDCl3, ppm): δ 137.2, 134.6, 125.3, 124.0, 91.4, 66.3, 29.2, 23.4,

21.9, 13.5, 12.5. HRMS (EI) m/z calc. for C11H18O: 166.1358, found 166.1353.

(S,Z)-8-((E)-But-2-en-2-yl)-7-methyl-3,4,5,8-tetrahydro-2H-oxocine (38). Following general procedure B, 24 (14 mg, 0.061 mmol), 16a (3 mg, 0.003 mmol), and NaI (9 mg,

0.06 mmol) in 1.1 mL THF afforded 38 as only 5% of a mixture of unreacted 24 and other olefinic products in 88% ee. Chiraldex G-TA, 1mL/min, 60 °C for 90 min, retention times = 57.4 (minor) and 58.7 (major) min.

(S,E)-6-(But-2-en-2-yl)-2,2,5-trimethyl-3,6-dihydro-2H-1,2-oxasiline (39). Following general procedure A, 26 (0.95 g, 4.0 mmol) and 18a (35 mg, 0.032 mmol) in 72 mL

CH2Cl2 gave 0.60 g (77% yield) of 39 as a yellow oil (3% EtOAc in hexanes) in 92% ee.

Chiraldex G-TA, 1mL/min, 60 °C for 60 min, retention times = 28.6 (minor) and 29.9

25 1 (major) min. [α]D = +195.4 (CHCl3, c = 0.96). H NMR (300 MHz, CDCl3, ppm): δ

5.69 (dquint, J = 7.7, 1.4 Hz, 1H), 5.49 (q, J = 6.6 Hz, 1H), 4.54 (s, 1H), 1.63 (dd, J = 6.6,

1.1 Hz, 3H), 1.54 (t, J = 1.1 Hz, 3H), 1.51 (s, 3H), 1.29–1.39 (m, 1H), 1.12–1.21 (m, 1H),

13 0.19 (s, 3H), 0.13 (s, 3H). C NMR (75 MHz, CDCl3, ppm): δ 136.9, 136.0, 122.9,

120.4, 83.4, 22.0, 13.5, 12.5, 10.7, 0.3, –0.6. HRMS (EI) m/z calc. for C11H20OSi:

196.1284, found 196.1281.

(S,Z)-7-((E)-But-2-en-2-yl)-3,3,6-trimethyl-2,3,4,7-tetrahydro-1,3-oxasilepine (40).

Following general procedure A, 25 (40 mg, 0.16 mmol) and 18a (1.7 mg, 0.0016 mmol)

in 2.9 mL CH2Cl2 gave 21.7 mg (65% yield) of 40 as a yellow oil (2% EtOAc in hexanes)

99 in 92% ee. Chiraldex G-TA, 1mL/min, 60 °C for 60 min, retention times = 28.7 (minor)

25 1 and 29.8 (major) min. [α]D = +184.3 (CHCl3, c = 0.75). H NMR (300 MHz, CDCl3, ppm): δ 5.66 (t, J = 7.4 Hz, 1H), 5.49 (q, J = 6.9 Hz, 1H), 4.49 (s, 1H), 2.55–2.67 (m,

1H), 2.02–2.12 (m, 1H), 1.68 (t, J = 1.1 Hz, 3H), 1.64 (d, J = 6.9 Hz, 3H), 1.56 (s, 3H),

13 0.75–0.86 (m, 2H), 0.16 (s, 3H), 0.14 (s, 3H). C NMR (75 MHz, CDCl3, ppm): δ 137.2,

136.5, 128.8, 121.8, 84.0, 22.6, 21.9, 16.8, 13.5, 11.8, 0.9, –0.3. HRMS (EI) m/z calc. for

C12H22OSi: 210.1440, found 210.1449.

(S,Z)-2,2,5-Trimethyl-7-((E)-2-methylbut-2-enyl)-2,3,6,7-tetrahydro-1,2-oxasilepine

(44). Following general procedure B, 35 (40 mg, 0.15 mmol), 8a (6.0 mg, 0.006 mmol), and NaI (23 mg, 0.15 mmol) in 2.7 mL THF gave 33.1 mg (98% yield) of 44 as a light yellow oil (2% EtOAc in hexanes) in 78% ee with <2% of 45 (diagnostic peaks at δ 4.78

(br s, 1H) and 4.72 (br s, 1H) in the 1H NMR spectrum). CP Chirasil-Dex-CB, 1 mL/min,

24 60 °C for 250 min, retention times = 205.5 (major) and 213.8 (minor) min. [α]D = +8.3

1 (CHCl3, c = 0.99). H NMR (300 MHz, CDCl3, ppm): δ 5.53 (t, J = 7.4 Hz, 1H), 5.24

(qq, J = 6.6, 1.1 Hz, 1H), 3.97–4.05 (m, 1H), 2.21–2.36 (m, 2H), 2.01–2.10 (m, 2H), 1.70

(s, 3H), 1.62 (s, 3H), 1.59 (d, J = 6.6 Hz, 3H), 1.54–1.60 (m, 1H), 1.31–1.39 (m, 1H),

13 0.11 (s, 3H), 0.08 (s, 3H). C NMR (75 MHz, CDCl3, ppm): δ 134.3, 133.4, 121.3,

121.2, 71.1, 49.2, 41.3, 25.9, 18.0, 16.2, 13.6, 0.4, –1.4. HRMS (EI) m/z calc. for

C13H24OSi: 224.1597, found 224.1598.

(S,E)-6-(But-2-en-2-yl)-5-methyl-2,2-diphenyl-3,6-dihydro-2H-1,2-oxasiline (41).

Following general procedure A, 27 (25 mg, 0.069 mmol) and 18a (1.5 mg, 0.0014 mmol)

100 in 1.3 mL CH2Cl2 gave crude 41 as a pale yellow oil (5% Et2O in pentane) in 80% ee.

Chiral SFC (supercritical CO2 with 5%–50% MeOH ramp over 10 min), ADH,

4 mL/min, 100 bar, 214 nm detector wavelength, retention times = 1.44 (major) and 1.99

1 (minor) min. H NMR (300 MHz, CDCl3, ppm): δ 7.57–7.64 (m, 4H), 7.32–7.43 (m,

6H), 5.81–5.85 (m, 1H), 5.55 (q, J = 6.6 Hz, 1H), 4.76 (br s, 1H), 1.84 (dq, J = 17.3, 2.8

Hz, 1H), 1.69–1.78 (m, 1H), 1.64 (d, J = 6.6 Hz, 3H), 1.63 (s, 3H), 1.53 (s, 3H). 13C

NMR (75 MHz, CDCl3, ppm): δ 137.02, 136.62, 136.08, 135.83, 134.73, 134.59, 130.20,

130.07, 128.09, 128.04, 123.32, 120.02, 84.27, 22.20, 13.59, 11.03, 10.50. HRMS (EI)

+ m/z calc. for C21H24OSi [M ] 320.1596, found 320.1597.

(E)-4-Methyl-2-(2-methylbut-2-enyl)-3,6-dihydro-2H-pyran (42). Following general procedure B, 34 (23 mg, 0.11 mmol), 8a (5 mg, 0.005 mmol), and NaI (17 mg, 0.11 mmol) in 2.0 mL THF afforded a crude mixture of 42 (>95%) in 35% ee and 43 (<5%) in

38% ee (diagnostic peaks at δ 4.78 (br s, 1H) and 4.82 (br s, 1H) in the 1H NMR spectrum). Chiraldex G-TA, 1mL/min, 60 °C for 70 min, retention times = 60.6 (minor) and 62.4 (major) min for 42, and 25.4 (minor) and 26.2 (major) min for 43. 1H NMR

(300 MHz, CDCl3, ppm) for 42: δ 5.39 (br s, 1H), 5.26–5.30 (m, 1H), 4.11–4.14 (m, 2H),

3.58–3.67 (m, 1H), 2.32 (dd, J = 13.7, 6.9 Hz, 1H), 2.11 (dd, J = 13.7, 6.3 Hz, 1H), 1.73–

1.90 (m, 2H), 1.68 (s, 3H), 1.63 (d, J = 0.8 Hz, 3H), 1.59 (dd, J = 6.9, 0.8 Hz, 3H).

Allyl((2E,5E)-hepta-2,5-dien-4-yloxy)dimethylsilane (48). To a solution of imidazole

(486 mg, 7.1 mmol) in DMF (5 mL) was added the known alcohol (2E,5E)-hepta-2,5- dien-4-ol27 (200 mg, 1.8 mmol) and allylchlorodimethylsilane (0.40 mL, 2.7 mmol).

101 After 16 h at rt, the reaction mixture was diluted with 10 mL of water and extracted with

Et2O (3 × 15 mL). The combined organic layers were washed with aqueous 1 M CuSO4,

water, dried over MgSO4, and concentrated. Purification by flash chromatography (2% ethyl acetate in hexanes) afforded 253 mg (67% yield) of 48 as a colorless oil. 1H NMR

(300 MHz, CDCl3, ppm): δ 5.72–5.86 (m, 1H), 5.54–5.65 (m, 2H), 5.42–5.50 (m, 2H),

4.83–4.92 (m, 2H), 4.52 (tquint, J = 6.0, 0.8 Hz, 1H), 1.68 (dt, J = 6.0, 0.8 Hz, 6H), 1.62

(dt, J = 8.0, 1.1 Hz, 2H), 0.11 (s, 6H). Upon exposure to olefin metathesis catalyst 50, it was completely converted into a complex mixture of volatile products.

((2E,5E)-3,5-Dimethylhepta-2,5-dien-4-yloxy)dimethyl(vinyl)silane (51). To a

solution of 21 (0.50 g, 3.6 mmol) in CH2Cl2 (10 mL) at 0 °C was added triethylamine (1.5 mL, 11 mmol) and chlorodimethylvinylsilane (0.98 mL, 7.1 mmol). After 12 h at rt, the light brown reaction mixture was poured into 20 mL of water and was extracted with

CH2Cl2 (3 × 30 mL). The combined organics were washed with water, brine, dried over

Na2SO4, and concentrated. Purification by flash chromatography (100% hexanes)

1 afforded 0.70 g (88% yield) of 51 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ

6.11 (dd, J = 20.1, 14.8 Hz, 1H), 5.95 (dd, J = 14.8, 4.4 Hz, 1H), 5.73 (dd, J = 20.1, 4.4

Hz, 1H), 5.52 (qquint, J = 6.6, 1.4 Hz, 2H), 4.30 (br s, 1H), 1.61 (dt, J = 6.6, 1.1 Hz, 6H),

1.42 (t, J = 1.1 Hz, 6H), 0.14 (s, 6H). Upon exposure to olefin metathesis catalyst 50, only a small amount was converted into the desired product 52; most of the 51 did not react.

102 (2E,5E)-3,5-Dimethylhepta-2,5-dien-4-yl acrylate (53). To a solution of N,N- dimethylaminopyridine (87 mg, 0.71 mmol), 21 (200 mg, 1.4 mmoml), and acrylic acid

(0.29 mL, 4.3 mmol) in CH2Cl2 at 0 °C was added 1-[3-(dimethylamino)propyl]-3- ethylcarbodiimide hydrochloride (EDC) (820 mg, 4.3 mmol) and triethylamine (0.60 mL,

4.3 mmol). After 15 min at 0 °C, the orange mixture was allowed to warm to rt, where it stirred for 3 days. It was diluted with 25 mL diethyl ether and 25 mL water, and was extracted with ether (3 × 30 mL). The combined organic layers were washed with

aqueous 1 N HCl, water, brine, dried over MgSO4, and concentrated. Purification by flash chromatography (4% ethyl acetate in hexanes) afforded 152 mg (55% yield) of 53

1 as a colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 6.42 (dd, J = 17.3, 1.6 Hz, 1H),

6.15 (dd, J = 17.3, 10.4 Hz, 1H), 5.82 (dd, J = 10.2, 1.6 Hz, 1H), 5.55 (tq, J = 6.6, 1.4 Hz,

2H), 5.51 (br s, 1H), 1.64 (dt, J = 6.6, 1.1 Hz, 6H), 1.54 (t, J = 1.1 Hz, 6H). Upon exposure to olefin metathesis catalyst 50, it was completely converted to a complex mixture of unseparated products that had many vinylic hydrogen atoms.

1. Cp2TiCl2 (3 mol %)

MgBr OH

2. ethyl formate Et2O, 0 °C to rt 68% yield 65

(3E,6E)-4,6-Diethylnona-3,6-dien-5-ol (65). To a solution of 3-hexyne (1.4 mL, 1.0 g,

12.2 mmol) and isobutylmagnesium bromide (2.0 M in Et2O, 6.1 mL, 12.2 mmol) in 12 mL of Et2O was added titanocene dichloride (85 mg, 0.34 mmol). After 1 h at rt, the solution was cooled to 0 °C and ethyl formate (0.44 mL, 0.40 g, 5.5 mmol), in 0.5 mL

Et2O, was added dropwise. After 5 min at 0 °C, the solution was allowed to warm to rt.

103 After 1 h at rt, the reaction was quenched with saturated aqueous ammonium chloride

(carefully; bubbling occurred), and it was filtered through a pad of celite, which was washed with water and diethyl ether. The organic layer was removed from the filtrate,

and the remaining aqueous layer was extracted with 2 × Et2O. The combined organic

layers were dried over Na2SO4 and concentrated. Purification by flash chromatography

(10% ethyl acetate in hexanes) afforded 0.73 g (68% yield) of 65 as a yellow oil. 1H

NMR (300 MHz, CDCl3, ppm): δ 5.47 (t, J = 7.2 Hz, 2H), 4.45 (br s, 1H), 1.98–2.13 (m,

6H), 1.81–1.93 (m, 2H), 1.44 (br s, 1H), 1.00 (t, J = 7.4 Hz, 6H), 0.94 (t, J = 7.4 Hz, 6H).

Br OH O NaH

THF, 65 °C 93% yield 65 55

(3E,6E)-5-(Allyloxy)-4,6-diethylnona-3,6-diene (55). To a suspension of 95% NaH (24 mg, 1.0 mmol) in 1.5 mL of THF was added 65 (100 mg, 0.51 mmol). After 10 min at rt, allyl bromide (filtered through neutral alumina, 66 µL, 0.76 mmol) was added. After 16

h at 65 °C, the reaction mixture was diluted with 10 mL Et2O, carefully quenched with 10 mL of water, and extracted with 3 × 20 mL of Et2O. The combined organic layers were

washed with brine (20 mL), dried over MgSO4, and concentrated. Purification by flash chromatography (1% ethyl acetate in hexanes) afforded 112 mg (93% yield) of 55 as a

1 colorless oil. H NMR (300 MHz, CDCl3, ppm): δ 5.86–5.99 (m, 1H), 5.45 (t, J = 7.2 Hz,

2H), 5.26 (dq, J = 17.3, 1.9 Hz, 1H), 5.10–5.16 (m, 1H), 4.00 (br s, 1H), 3.89 (dt, J = 5.5,

1.4 Hz, 2H), 1.96–2.14 (m, 6H), 1.78–1.90 (m, 2H), 0.99 (t, J = 7.4 Hz, 6H), 0.92 (t, J =

13 7.4 Hz, 6H). C NMR (75 MHz, CDCl3, ppm): δ 138.45, 135.67, 129.25, 116.20, 86.02,

104 + 68.92, 21.10, 20.61, 14.66, 14.20. HRMS (EI) m/z calc. for C16H28O [M ] 236.2140, found 236.2140.

(E)-3-Ethyl-2-(hex-3-en-3-yl)-2,5-dihydrofuran (56). Following general procedure B,

55 (12 mg, 0.050 mmol), 8a (2 mg, 0.0020 mmol), and NaI (7.5 mg, 0.050 mmol) in 0.9 mL THF gave a crude residue that was passed down a pipet column (5% ethyl acetate in hexanes) to afford 56 (>95% conv.) in 29% ee. Chiraldex G-TA, 1mL/min, 50 °C for 15

1 min, retention times = 2.7 (major) and 3.1 (minor) min. H NMR (300 MHz, CDCl3, ppm): δ 5.55–5.57 (m, 1H), 5.36 (t, J = 14.6 Hz, 1H), 4.95 (br s, 1H), 4.65–4.72 (m, 1H),

4.56–4.62 (m, 1H), 1.80–2.13 (m, 6H), 1.07 (t, J = 7.4 Hz, 3H), 0.99 (t, J = 7.4 Hz, 3H),

13 0.97 (t, J = 7.4 Hz, 3H). C NMR (75 MHz, CDCl3, ppm): δ 144.12, 140.19, 131.73,

119.40, 94.54, 75.51, 21.13, 20.36, 19.40, 15.01, 14.55, 12.03. HRMS (EI) m/z calc. for

+ C12H20O [M ] 180.1514, found 180.1510.

References

1 For recent reviews on asymmetric olefin metathesis, see (a) Hoveyda, A. H. In

Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH: Weinheim, Germany, 2003;

Chapter 2.3. (b) Schrock, R. R.; Hoveyda, A. H. Angew. Chem. Int. Ed. 2003, 42, 4592–

4633. (c) Hoveyda, A. H.; Schrock, R. R. Chem. Eur. J. 2001, 7, 945–950.

2 La, D. S.; Alexander, J. B.; Cefalo, D. R.; Graf, D. D.; Hoveyda, A. H.; Schrock, R. R.

J. Am. Chem. Soc. 1998, 120, 9720–9721.

3 Zhu, S. S.; Cefalo, D. R.; La, D. S.; Jamieson, J. Y.; Davis, W. M.; Hoveyda, A. H.;

Schrock, R. R. J. Am. Chem. Soc. 1999, 121, 8251–8259.

105

4 Seiders, T. J.; Ward, D. W.; Grubbs, R. H. Org. Lett. 2001, 3, 3225–3228.

5 (a) Van Veldhuizen, J. J.; Gillingham, D. G.; Garber, S. B.; Kataoka, O.; Hoveyda, A.

H. J. Am. Chem. Soc. 2003, 125, 12502–12508. (b) Van Veldhuizen, J. J.; Garber, S. B.;

Kingsbury, J. S.; Hoveyda, A. H. J. Am. Chem. Soc. 2002, 124, 4954–4955. (c) Van

Veldhuizen, J. J.; Campbell, J. E.; Giudici, R. E.; Hoveyda, A. H. J. Am. Chem. Soc.

2005, 127, 6877–6882.

6 (a) Gillingham, D. G.; Kataoka, O.; Garber, S. B.; Hoveyda, A. H. J. Am. Chem. Soc.

2004, 126, 12288–12290. (b) ref. 5.

7 The best result for ARCM with catalysts 9 and 10 is a 77% ee, and in many cases 10 mol % of the catalyst is needed.

8 Berlin, J. M.; Goldberg, S. D.; Grubbs, R. H. Submitted, 2006.

9 A portion of this work is published: Funk, T. W.; Berlin, J. M.; Grubbs, R. H. J. Am.

Chem. Soc. 2006, 128, 1840–1846.

10 Wolfe, J. P.; Buchwald, S. L. J. Org. Chem. 2000, 65, 1144–1157.

11 (a) Sato, F.; Ishikawa, H.; Sato, M. Tetrahedron Lett. 1981, 22, 85–88. (b) Garner, C.

M.; Prince, M. E. Tetrahedron Lett. 1994, 35, 2463–2464.

12 Han, S.-Y. and Chang, S. In Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH:

Weinheim, Germany, 2003; Chapter 2.2.

13 (a) Dias, E. L.; Nguyen, S. T.; Grubbs, R. H. J. Am. Chem. Soc. 1997, 119, 3887–3897.

(b) Sanford, M. S.; Love, J. A.; Grubbs, R. H. J. Am. Chem. Soc. 2001, 123, 6543–6554.

14 One mol % of 18a is added at the beginning of the reaction, and another 1 mol % is added after 2 h. When 2 mol % of 18a is added at the start with no subsequent additions, the yield was only 74%. For more information, see ref. 9.

106

15 See appendix 1 for an interesting side product formed during ARCM of 47.

16 (a) Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004, 126, 7414–7415.

(b) Ulman, M.; Grubbs, R. H. J. Org. Chem. 1999, 64, 7202–7207.

17 (a) Pietraszuk, C.; Marciniec, B.; Fischer, H. Organometallics 2000, 19, 913–917. (b)

Pietraszuk, C.; Fischer, H. Chem. Commun. 2000, 2463–2464. (c) Pietraszuk, C; Fischer,

H. Organometallics 2001, 20, 4641–4646.

18 Teng, X.; Cefalo, D. R.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2002, 124,

10779–10784.

19 Morgan, J. P. Ph.D. Thesis, California Institute of Technology, 2002.

20 Evidence for cis coordination: (a) Trnka, T. M.; Day, M. W.; Grubbs, R. H.

Organometallics 2001, 20, 3845–3847. Evidence for trans coordination: (b) Tallarico, J.

A.; Bonitatebus, Jr., P. J.; Snapper, M. L. J. Am. Chem. Soc. 1997, 119, 7157–7158. (c)

Romero, P. E.; Piers, W. E. J. Am. Chem. Soc. 2005, 127, 5032-5033.

21 (a) Costabile, C.; Cavallo, L. J. Am. Chem. Soc. 2004, 126, 9592–9600. (b) Adlhart, C.;

Chen, P. J. Am. Chem. Soc. 2004, 126, 3496–3510. (c) Fomine, S.; Vargas, S. M.;

Tlenkopatchev, M. A. Organometallics 2003, 22, 93–99. (d) Cavallo, L. J. Am. Chem.

Soc. 2002, 124, 8965–8973. (e) Vyboishchikov, S. F.; Bühl, M.; Thiel, W. Chem. Eur. J.

2002, 8, 3962–3975.

22 This explanation is directly related to the Curtin-Hammett principle. For more information, see Seeman, J. I. Chem. Rev. 1983, 83, 83–134.

23 Another explanation is that the reaction proceeds through a late, product-like transition state when the less active diiodide catalysts are used. A higher degree of stereochemical communication would be expected for a late transition state. For evidence of this affect

107 derived from electronic changes to a catalyst, see Palucki, M.; Finney, N. S.; Pospisil, P.

J.; Güler, M. L.; Ishida, T.; Jacobsen, E. N. J. Am. Chem. Soc. 1998, 120, 948–954.

24 Kiely, A. F.; Jernelius, J. A.; Schrock, R. R.; Hoveyda, A. H. J. Am. Chem. Soc. 2002,

124, 2868–2869.

25 Robertson, J.; Hall, M. J.; Stafford, P. M.; Green, S. P. Org. Biomol. Chem. 2003, 1,

3758–3767.

26 Kunai, A.; Kawakami, T.; Toyoda, E.; Ishikawa, M. Organometallics 1992, 11, 2708-

2711.

27 Underiner, T. L.; Goering, H. L. J. Org. Chem. 1990, 55, 2757–2761.